VEGF is a naturally occurring compound that is produced in follicular or folliculo-stellate cells (FC), a morphologically well characterized population of granular cells. The FC are stellate cells that send cytoplasmic processes between secretory cells.
Several years ago a heparin-binding endothelial cell-growth factor called vascular endothelial growth factor (VEGF) was identified and purified from media conditioned by bovine pituitary follicular or folliculo-stellate cells. See Ferrara et al., Biophys. Res. Comm. 161, 851 (1989).
Although a vascular endothelial cell growth factor could be isolated and purified from natural sources for subsequent therapeutic use, the relatively low concentrations of the protein in FC and the high cost, both in terms of effort and expense, of recovering VEGF proved commercially unavailing. Accordingly, further efforts were undertaken to clone and express VEGF via recombinant DNA techniques. The embodiments of that research are set forth in the patent applications referred to supra; this research was also reported in the scientific literature in Laboratory Investigation 72, 615 (1995), and the references cited therein.
In those applications there is described an isolated nucleic acid sequence comprising a sequence that encodes a vascular endothelial cell growth factor having a molecular weight of about 45,000 daltons under non-reducing conditions and about 23,000 under reducing conditions as measured by SDS-PAGE. Both the DNA and amino acid sequences are set forth in figures forming a part of the present application--see infra.
VEGF prepared as described in the patent applications cited supra, is useful for treating conditions in which a selected action on the vascular endothelial cells, in the absence of excessive tissue growth, is important, for example, diabetic ulcers and vascular injuries resulting from trauma such as subcutaneous wounds. Being a vascular (artery and venus) endothelial cell growth factor, VEGF restores cells that are damaged, a process referred to as vasculogenesis, and stimulates the formulation of new vessels, a process referred to as angiogenesis. VEGF is expressed in a variety of tissues as multiple homodimeric forms (121, 165, 189 and 206 amino acids per monomer) resulting from alternative RNA splicing. VEGF.sub.121 is a soluble mitogen that does not bind heparin; the longer forms of VEGF bind heparin with progressively higher affinity. The heparin-binding forms of VEGF can be cleaved in the carboxy terminus by plasmin to release (a) diffusible form(s) of VEGF. Amino acid sequencing of the carboxy terminal peptide identified after plasmin cleavage is Arg.sub.110 -Ala.sub.111. Amino terminal "core" protein, VEGF (1-110) isolated as a homodimer, binds neutralizing monoclonal antibodies (4.6.1 and 2E3) and soluble forms of fms-like tyrosine kinase (FLT-1), kinase domain region (KDR) and fetal liver kinase (FLK) receptors with similar affinity compared to the intact VEGF.sub.165 homodimer.
As noted, VEGF contains two domains that are responsible respectively for binding to the KDR and FLT-1 receptors. These receptors exist only on endothelial (vascular) cells. As cells become depleted in oxygen, because of trauma and the like, VEGF production increases in such cells which then bind to the respective receptors in order to signal ultimate biological effect. The signal then increases vascular permeability and the cells divide and expand to form new vascular pathways-vasculogenesis and angiogenesis. Thus, VEGF and derivatives thereof, as described in the patent applications referred to supra, would find use in the restoration of vasculature after a myocardial infarct, as well as other uses that can be deduced.
The present invention is predicated upon research intended to identify the regions or domains that are responsible for binding to the KDR and FLT receptors. After identification, it was a goal to mutagenize such a domain in order to produce variants that have either increased or decreased binding capability with respect to those respective KDR and FLT binding domains.
It was a further object of this research to produce VEGF variants that would have selective activity with respect to the binding KDR and FLT domains. It was postulated that if one could increase the binding capability of the domain responsible for vasculogenesis and angiogenesis, one could produce a more potent material for intended therapeutic use. Conversely, if one could by induced mutagenesis produce VEGF variants that had reduced activity, and consequently, anti-vasculogenesis and anti-angiogenesis, one could use such variants in instances of tumor treatment in order to starve the tumors for intended regression.
As further objects, such variants could then be employed in assay systems to discover small molecule agonists and antagonists for intended therapeutic use in such indications.
The results of such research is the subject of the present invention. The dominant domains of VEGF for receptor binding were found to be proximately located, but at distinct sites, allowing the development of variants that proved to be receptor-selective. The KDR receptor was found to bind VEGF predominantly through the sites on a putative loop which contains Arginine (Arg or R) at position 82 of VEGF, Lysine (Lys or K) at position 84 and Histidine (His or H) at position 86. The FLT-1 receptor was found to bind VEGF predominantly through the sites on a putative loop which contains Aspartic acid (Asp or D) at position 63, Glutamic acid (Glu or E) at position 64 and Glutamic acid (Glu or E) at position 67. Mutagenesis experiments followed with respect generally to these domains resulting in the variants of the present invention. Such mutagenesis employed both specific and random strategies, in accordance with procedures generally well known to the art-skilled.
Since VEGF functions by dimerizing or aggregating its target receptors an antagonist must knock out binding at just one of the KDR binding sites. In this way, an intact binding site remains and allows the hormone to bind receptor, but it is unable to activate KDR. Such a one to one complex (of mutant VEGF with a single KDR receptor) is functionally inert and prevents endogenous VEGF from interacting with KDR. Since VEGF is a homodimer with a two-fold symmetry, each mutation that disrupts binding at one site will necessarily identically disrupt binding at the other site. Consequently, to produce an antagonist it is necessary to develop a method of restricting mutations to just one site. This can in theory be accomplished by producing heterodimers or by making the VEGF homodimer into a single chain molecule.
Thus, produced are antagonists by making a single chain VEGF molecule. Since the two subunits become fused in a single molecule the number designation of the residues in what would be equivalent to the 2nd subunit is changed.